polymer jamming using Stillinger-Lubachevsky growth algorithm and time-dependent LAMMPS variables?

LAMMPS LAMMPS Mailing List Mirror
1

Hi, everybody. I want to simulate jamming of semiflexible polymers using a Stillinger-Lubachevsky-like growth algorithm. The monomers will have some short-ranged, purely repulsive pair potential (e.g. pair_style gran/hooke), harmonic bonds (bond_style harmonic), and harmonic angular interactions (angle_style harmonic).

The question I have is how to implement the S-L particle-growth algorithm for this potential. In S-L, the particle diameters start out small, then increase with time at some defined rate. Can this be implemented in LAMMPS scripts using some sequence of commands involving “variables”? For example, by defining a variable corresponding to particle diameter and then telling LAMMPS it should increase at some defined rate? I think this can be done in stock LAMMPS, perhaps using “atom”-style variables for particle radius and mass, but am not sure how to do this.

Also, one potentially tricky issue for simulating polymer jamming is that the bond and angle interactions also need to “know” about the increasing monomer (particle) diameter. Can LAMMPS do this in its current form?

As always, any advice will me much appreciated :).

Thanks,

Rob

2

Hi, everybody. I want to simulate jamming of semiflexible polymers using a
Stillinger-Lubachevsky-like growth algorithm. The monomers will have some
short-ranged, purely repulsive pair potential (e.g. pair_style gran/hooke),
harmonic bonds (bond_style harmonic), and harmonic angular interactions
(angle_style harmonic).

The question I have is how to implement the S-L particle-growth algorithm
for this potential. In S-L, the particle diameters start out small, then
increase with time at some defined rate. Can this be implemented in LAMMPS
scripts using some sequence of commands involving “variables”? For example,
by defining a variable corresponding to particle diameter and then telling
LAMMPS it should increase at some defined rate? I think this can be done in
stock LAMMPS, perhaps using “atom”-style variables for particle radius and
mass, but am not sure how to do this.

please have a look at fix adapt. this is meant to be the generic
mechanism to change simulation parameters over time. it may be needed
to add corresponding interfaces to specific styles, if needed, but the
procedure is usually rather straightforward.

Also, one potentially tricky issue for simulating _polymer_ jamming is that
the bond and angle interactions also need to “know” about the increasing
monomer (particle) diameter. Can LAMMPS do this in its current form?

again, it should be possible to set this up via fix adapt. the logic
of keeping changes in sync would have to be implemented the variable
expressions.

axel

3

The fix adapt command will give you what you want when the interaction
diameter is stored in the pair_style e.g.

fix 1 all adapt 1 pair lj/cut sigma * * v_scale

If you want to change the diameter of a granular particle, the doc
page says this works:

variable size equal ramp(1.0,0.1)
fix 1 center adapt 10 atom diameter v_size

I don't know what "know" means, so I can't address the question of the
bond and angle.

Aidan

2015-11-13 4:37 GMT-07:00 Robert Hoy <[email protected]...>:

4

If a couple hooks are added to fix adapt, it should
also be possible to vary a bond potential’s parameters
over time, e.g. the equilibrium bond length. In that
case the bond length could grow as the particle
diameter increased. Don’t know what the analog
is for varying an angle.

Steve